More than 90% of breast cancers arise from epithelial cells or from progenitors with stem cell properties. Cell polarity is essential for stem cell asymmetric divisions, and failures in polarization can result in tumorigenesis. The Par polarity proteins are key regulators of asymmetric cell division. Par proteins also control polarization of differentiated epithelial cells. Yet the molecular mechanisms underlying polarity in mammary morphogenesis and the role of Par proteins in breast cancer remain largely unknown. We have developed methods that facilitate the study of mammary morphogenesis and tumorigenesis. To manipulate gene expression in mammary glands, stem cells are transduced with lentivirus, and mammary glands are regenerated in host mice from these cells. Lentiviral transduction enables rapid analysis of molecular mechanism through a knock-down/knock-in approach. Using these methods, an essential role for the Par3 polarity protein in mammary morphogenesis was discovered. In addition, silencing of Par3 potentiated tumor growth and metastasis. The goals are as follows: 1. How does Par3 regulate stem/progenitor self-renewal and differentiation during mammary gland morphogenesis? In the absence of Par3, atypical protein kinase C (aPKC) is mislocalized from the luminal apical surface. A knock-down/knock-in strategy will be used to ask if fate determination can be rescued by apically-tethered aPKC and by Par3 mutants. A novel miRNA sensor will be used to identify stem cells and to ask if Par3 is required for asymmetric division. Differential functions of Par3 in luminal and myoepithelial cells will be tested using new methods for cell-type specific RNAi silencing. 2. How does Par3 function to suppress tumor growth and metastasis? Preliminary data suggest that Par3 might limit mammary cell proliferation by suppressing Rac activity in the NICD cells, while separately controlling cell lineage determination and metastasis through aPKC. Lentivirus will be used to manipulate gene expression in primary cells, which are tested using the in vivo transplant model, tail vein injections, and in vitro mammosphere cultures. 3. Can we identify multiple additional metastasis suppressors/inducers by in vivo screening?. Mammary stem cells from ErbB2 mice will be infected with pooled lentiviruses that express a library of cDNAs, then implanted into cleared fat pads of wild type host mice. The viruses also express a red fluorescent protein. RFP-positive lung metastases will be isolated and screened by PCR to identify the cDNA associated with each metastatic colony. Positives will be prioritized based on expression patterns in human tumors for mechanistic studies. Together, these studies will provide insights into polarity protein signaling during stem/progenitor cell self-renewal and lineage determination, and identify new players in mammary tumor metastasis.